参考资料

• https://www.bilibili.com/video/av27122140?from=search&seid=9460780308750646835
• 笔记完整版(转载)
• https://docker.com
• https://docker-cn.com
• https://hub.docker.com/

目标检测：

https://blog.csdn.net/electech6/article/details/95240278

RPN：

https://blog.csdn.net/shiheyingzhe/article/details/83713352

YOLO
https://blog.csdn.net/shuiyixin/article/details/82533849
https://luckmoonlight.github.io/2018/11/28/yoloV1yolov2yoloV3/

YOLO v2
https://blog.csdn.net/weixin_40092412/article/details/91547836

refinet
https://blog.csdn.net/zhangjunhit/article/details/72844862
https://github.com/DrSleep/refinenet-pytorch/blob/master/models/resnet.py

1. 前言

CVPR2019 看到好几篇都使用了 memory 的结构来进行 person re-id，所以对这几篇论文总结一下。具体的论文分析在我之前的博客已经有，这里主要介绍这几篇论文对 memory 的使用方法和效果。

• OIM: CVPR2017_Joint Detection and Identification Feature Learning for Person Search
• DIMN: CVPR2019_Generalizable Person Re-identification by Domain-Invariant Mapping Network
• MAR: CVPR2019 Unsupervised Person Re-identification by Soft Multilabel Learning
• ECN: CVPR2019_Invariance Matters Exemplar Memory for Domain Adaptive Person Re-identification

1. Introduction

• paper: CVPR2017_Joint Detection and Identification Feature Learning for Person Search
• code: caffe, [pytorch][https://github.com/Cysu/open-reid]
• project: End-to-End Deep Learning for Person Search
• memory: OIM, DIMN, MAR, ECN

1. 前言

• paper: CVPR2019_Generalizable Person Re-identification by Domain-Invariant Mapping Network
• code: 暂无

0. 前言

• paper: CVPR2019_ArcFace: Additive Angular Margin Loss for Deep Face Recognition
• code: MXNet, pytorch, pytorch, tensorflow
• author: 邓健康

开源代码中称之为 insightface，思路简单，效果却很好，并且与其他变种做了详尽的对比。(这篇论文原名是ArcFace，但是由于与虹软重名，后改名为Insight Face)

0. 前言

• paper: FUNIT: Few-Shot Unsupervised Image-to-Image Translation
• code: code

代码还需要等段时间，才能下载到所有代码。

这篇文章和郑哲东的 Joint Discriminative and Generative Learning for Person Re-identification 这篇文章看着有点像，果然是高手的思路都是相同的。不对，两个都是 NVIDIA 的，哈哈哈。

这篇文章思路奇特，实现简单。

0. 前言

• paper: CVPR2019_STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing
• code: tensorflow, pytorch

0. 前言

• paper: CVPR2019 Weakly Supervised Person Re-Identification
• code: